PDF Archive

Easily share your PDF documents with your contacts, on the Web and Social Networks.

Share a file Manage my documents Convert Recover PDF Search Help Contact



nclimate2552 .pdf



Original filename: nclimate2552.pdf
Title: Near-term acceleration in the rate of temperature change
Author: Steven J. Smith; James Edmonds; Corinne A. Hartin; Anupriya Mundra; Katherine Calvin

This PDF 1.5 document has been generated by LaTeX with hyperref package / pdfTeX-1.40.14, and has been sent on pdf-archive.com on 11/03/2015 at 06:07, from IP address 1.120.x.x. The current document download page has been viewed 827 times.
File size: 760 KB (4 pages).
Privacy: public file




Download original PDF file









Document preview


LETTERS
PUBLISHED ONLINE: 9 MARCH 2015 | DOI: 10.1038/NCLIMATE2552

Near-term acceleration in the rate of
temperature change
Steven J. Smith*, James Edmonds, Corinne A. Hartin, Anupriya Mundra and Katherine Calvin
Anthropogenically driven climate changes, which are expected
to impact human and natural systems, are often expressed in
terms of global-mean temperature1 . The rate of climate change
over multi-decadal scales is also important, with faster rates
of change resulting in less time for human and natural systems
to adapt2 . We find that present trends in greenhouse-gas and
aerosol emissions are now moving the Earth system into a
regime in terms of multi-decadal rates of change that are
unprecedented for at least the past 1,000 years. The rate of
global-mean temperature increase in the CMIP5 (ref. 3) archive
over 40-year periods increases to 0.25 ± 0.05 ◦ C (1σ) per
decade by 2020, an average greater than peak rates of change
during the previous one to two millennia. Regional rates of
change in Europe, North America and the Arctic are higher than
the global average. Research on the impacts of such near-term
rates of change is urgently needed.
Global-mean surface temperatures (GMT) are a useful indicator
of the overall scale of anthropogenic climate change over time,
although changes in other variables, such as local temperatures,
seasonal and diurnal temperature patterns, precipitation and storm
tracks, are of more direct relevance to climate impacts1 . Although
rates of climate change are implicit in all climate projections,
and occasionally mentioned relative to a century timescale, subcentury rates of change have rarely been examined. O’Neill and
Oppenheimer2 stressed the importance of the rate of climate change,
although we find a more constrained range of near-term rates in this
updated analysis (see Supplementary Section 7). Recently Ross et al.
examined the sensitivity of the rate of climate change to ocean heat
uptake, but only as averages over the entire twenty-first century4 .
We focus here on rates of change over 40-year periods, comparing
past trends with future projections. These represent background
trends averaging over shorter-term internal variability. Over this
timescale rates from long palaeo-climate records begin to become more reliable (or at least more consistent with model results, Supplementary Section 3.1), historical changes to date are
becoming distinguishable from modelled variability (Supplementary Section 3.5), and this begins to be comparable to the lifetime of much of human infrastructure. Rates of change from a
number of temperature reconstructions for Northern Hemisphere
areas are shown in Fig. 1. Forty-year warming trends rarely show
an average rate very much above 0.1 ◦ C/decade for the 900 years
before the twentieth century, with some reconstructions showing
occasional periods with trends up to 0.2 ◦ C/decade. Millennial-scale
coupled carbon–climate model simulations show similar overall
behaviour5 (Supplementary Fig. 4). During the latter half of the
twentieth century anthropogenic influences increase in importance,
with the Northern Hemisphere rate of climate change now above
0.2 ◦ C/decade. Rates of change can be larger when considered over a

shorter 20-year time period, but recent rates of change are not easily
distinguishable from natural variability on this timescale (Supplementary Figs 2 and 14).
Although global temperature trends are one of the most
commonly used metrics of climate change, climate impacts will
be driven by regional trends. Frequency distributions for regional
trends before the period of strong anthropogenic influence are
shown in Fig. 2 from two sources: reconstruction extending over the
past 2,000 years from the PAGES 2k consortium6 and the CMIP5 set
of climate model runs3 . Because millennial-scale model runs are not
available for most climate models, we instead conduct an analysis
over a shorter historical time period (1850–1930), considering
one ensemble member from each model in the CMIP5 (Coupled
Model Intercomparison Project phase 5) archive with consistent
data (see Supplementary Information). This ‘model-time’ domain
analysis samples over a range of internal variability and model
responses, although it considers only the recent history for solar and
volcanic forcings.
Both the CMIP5 and PAGES 2k analysis show distinct regional
differences in the rate of change frequency distribution (Fig. 2). A
number of regions show changes greater than 0.1 ◦ C per decade in
both data sets. Australasia has lower rates of change in both data
sets. On average, the CMIP5 data have a 20–30% larger range than
the PAGES 2k data. These differences could be due to biases in
either data set, but do not alter our conclusions (see Supplementary
Section 3.1).
Future rates of temperature change are examined using consistent
projections of future emissions of greenhouse gases, aerosol and
aerosol precursors, and other criteria pollutant emissions. Figure 3a
shows the 5–95% range for rates of change in the CMIP5 archive
over all 40-year periods ending between 2011 and 2020 under the
RCP4.5 scenario compared with a range of rates from a historical
period where anthropogenic forcing is weak. By showing the range
of trends across CMIP5 models we explicitly consider the impact
of internal variability over this period, at least to the extent this
is captured in the models. Figure 3b shows trends for the average
CMIP5 regional 40-year rate of change.
By 2020, both the range in the rate of change, and the average
value, is substantially higher than in the historical period (Fig. 3).
For most regions, there is only a small overlap between the lowest
5% of CMIP5 rates of change in this decade and the 95% point
from the 1891–1910 period (see also Supplementary Section 3.2).
This implies that there is a relatively small chance that 40-year
temperature trends over the 2011–2020 decade for 40-year periods
will be within pre-industrial bounds. We find that rates of change,
relative to inter-regional differences, generally do not have a strong
dependence on the size of the analysis region or the ocean fraction
(Supplementary Fig. 10).

Joint Global Change Research Institute, Pacific Northwest National Laboratory, 5825 University Research Court, Suite 3500, College Park, Maryland
20740, USA. *e-mail: ssmith@pnnl.gov
NATURE CLIMATE CHANGE | ADVANCE ONLINE PUBLICATION | www.nature.com/natureclimatechange
© 2015 Macmillan Publishers Limited. All rights reserved.

1

NATURE CLIMATE CHANGE DOI: 10.1038/NCLIMATE2552

LETTERS
0.3

Temperature trend (°C per decade)

0.2

‘Northern Hemisphere’ rate of change (40-year trends)
MNN2008

MJ2003

BOS..2001

B2000

JBB..1998

ECS2002

RMO..2005

MSH..2005

DWJ2006

HCA..2006

O2005

PS2004

CL2011_g

HADCRU4

0.1

0.0

−0.1

−0.2

−0.3
1000

1200

1400

1600

1800

2000

Year

0.2
0.1
0.0
−0.1
−0.2
−0.3

Europe

Asia

South
America

Australasia

Arctic

Antarctica

Figure 2 | 40-year rates of change from the PAGES 2k reconstructions up
to 1900 and the CMIP5 climate model archive for the period 1850–1930.
The bars (solid—PAGES 2k reconstruction, dotted—CMIP5 data) show the
5–95% range of occurrence for rates of change within each data set. The
CMIP5 data were processed to match the seasonal and spatial (land +
ocean or land only) coverage of the PAGES 2k reconstruction target for
each region (Supplementary Table 2).

This is consistent with findings that recent temperature extremes
in high northern latitudes are unprecedented when compared
with the past six centuries7 . We note that Antarctica stands out
with a lower rate of change signal relative to the inter-model
variation, indicating less agreement for modelled trends in this
region (Supplementary Fig. 6).
A simple climate model is used to diagnose the primary drivers
of the forced component of these rates of change. In each future
emissions scenario considered, we examine the impact of uncertainties in the equilibrium climate sensitivity and the assumed
strength of aerosol forcing in the absence of internal variability by
varying the values for each of these parameters (see Supplementary Information). Climate sensitivity is varied between 1.5 and
4.5 ◦ C/CO2 -doubling, which a likely range of climate sensitivity
from the IPCC AR5 (ref. 8). We vary total aerosol forcing between
−0.4 and −1.4 W m−2 in 2000, which we judge to cover a comparable probability range. All combinations of parameters are examined
to explore the range. Aerosols at present provide a net cooling effect
2

a
Rate of change (°C per decade)

0.3

90% occurrence range
PAGES 2k and CMIP5 (1850–1930) season, land/ocean match

b
Rate of change (°C per decade)

Rate of change (°C per decade)

Figure 1 | Rates of temperature change over 40-year periods for a number of climate reconstructions that cover various Northern Hemisphere areas.
Areas are land and ocean, land, and land 20◦ –90◦ , see Supplementary Information. The thick black line shows the corresponding rate of change of
Northern Hemisphere temperature from the HADCRU4 data set13 . Trends are linear fits ending at the year shown. Note that the proxy and historical
records over the twentieth century are generally not independent, because proxy records are often calibrated to the historical temperature record. Global
rates of change are about 20% smaller than Northern Hemisphere land + ocean rates of change in the CMIP5 archive (Supplementary Section 3.4).
Range for 40-year rates of change
CMIP5 (1851–1930 versus 1971–2020)

0.80
0.60
0.40
0.20
0.00
−0.20
−0.40

Europe

0.7
0.6
0.5
0.4
0.3
0.2

Asia

North
South Australasia Arctic Antarctica
America America

Average 40-year rate of change: CMIP5 RCP4.5 scenario
Global
Europe
Asia
South America
Australasia
Arctic
Antarctica
North America

0.1
0.0
−0.1
1900 1920 1940 1960 1980 2000 2020 2040 2060 2080 2100
Year

Figure 3 | Past and future regional rates of change from CMIP5. a, 5–95%
occurrence range of the decadal rate of change over 40-year periods from
the CMIP5 archive over a time with small anthropogenic influences
(1851–1930, dotted lines) and a period centred on the present (40-year
periods ending in 2011–2020, solid lines) from the RCP4.5 scenario.
Because the CMIP5 archive is a ‘sample of convenience’, these percentages
should not be interpreted as probabilities. b, Average over CMIP5 models
for the regional rate of change. Rates of change in this figure are annual
averages over land + ocean areas in each region.

and the imposition of pollution controls around the world have been
reducing emissions, particularly of sulphur dioxide9,10 , resulting in a
net warming.
Figure 4 shows the 40-year global rate of change for two future
scenarios used to drive the CMIP5 climate models, RCP4.5 and

NATURE CLIMATE CHANGE | ADVANCE ONLINE PUBLICATION | www.nature.com/natureclimatechange
© 2015 Macmillan Publishers Limited. All rights reserved.

NATURE CLIMATE CHANGE DOI: 10.1038/NCLIMATE2552
a

Global 40 year rate of change—RCP8.5

Rate of change (°C per decade)

Aerosol forcing
Climate sensitivity
0.4

Central case
Maximum
Minimum
Observed

0.2

0.0

1950

2000

2050

2100

Year

b

Global 40 year rate of change—RCP4.5

Rate of change (°C per decade)

Aerosol forcing
Climate sensitivity
0.4

Central case
Maximum
Minimum
Observed

0.2

0.0

1950

2000

2050

2100

Year

Figure 4 | Decomposition of global rates of temperature change from the
MAGICC model. a,b, Forty-year rates of change for periods the RCP 8.5
scenario (a) and the GCAM 4.5 W m−2 climate stabilization scenario (b).
Results are shown for: central climate assumptions (thick solid line), range
due to uncertainty in aerosol forcing (grey shading), and range due to
uncertainty in climate sensitivity (blue shading). The outer bounding cases
are shown as dotted lines. The thin solid black line shows the historical rate
of change using the HADCRU4 observational data. The vertical dashed line
indicates 2014.

RCP8.5. The RCP4.5 scenario represents a world where global
actions are taken to limit greenhouse gas concentrations11 , whereas
the RCP8.5 examines a contrasting case where greenhouse gas
emissions remain at high levels throughout the century12 . Also
shown, for reference, is the historical rate of change derived from
the HADCRU4 record13 . By 2030, the 40-year rate of change spans
0.15 to 0.33 ◦ C/decade. The range in future rates of change from the
simple model are comparable to the CMIP5 range (Supplementary
Fig. 17). A number of the CMIP5 models, however, do not reproduce
the historical drop in the rate of change seen in both the simple
model and the HADCRU record.
Aerosol forcing assumptions strongly impact the rate of change
until near the end of the twentieth century, but have a much smaller
relative impact over the twenty-first century. By 2010, most of the

LETTERS
spread in results is due to the assumed range in climate sensitivity.
The spread in results in the twenty-first century is dominated by
the assumed climate sensitivity, with a smaller impact for aerosol
assumptions. The near-term rates of change are not very sensitive to
the emissions scenario, with almost no difference out to 2025 (Fig. 4
and Supplementary Fig. 16), and successively larger differences after
this point.
During the historical period, a range of assumptions impact
rates of change, including internal (unforced) variability, solar and
volcanic forcings, and anthropogenic forcings, particularly those
for aerosols. Aerosol forcing has a particularly large impact in the
twentieth century because these occur predominantly over land,
which has a faster response to forcing than forcing over oceans.
The above analyses show that the world is at present in the midst
of a transition to a regime where the rate of climate change on
both global and regional scales will be dominated by the impact
of increasing greenhouse gas forcing14 . In this regime, assumptions
for the magnitude of the climate sensitivity, which is still poorly
constrained, has a dominant effect on rate of climate change over
those of aerosol forcing assumptions (Fig. 4).
A climate policy that limits total radiative forcing to 4.5 W m−2
by the end of the century (Fig. 4b) limits the potential rate of
change in the long term, but does not appreciably alter global-mean
temperatures over the next few decades. This is due to both the
impact of ocean thermal inertia and the fact that net aerosol cooling
declines in the near-term as climate policy is implemented15,16 .
Under the 4.5 W m−2 stabilization scenario, rates of global-mean
temperature change are still above 0.2 ◦ C/decade until after the
mid twenty-first century unless the climate sensitivity is low (for
example, ≤2.0 ◦ C/CO2 -doubling).
If the climate sensitivity is as low as 1.5 ◦ C/CO2 -doubling, then
the rate of change stabilizes around 0.15 ◦ C/decade. In contrast, if
the climate sensitivity is as high as 4.5 ◦ C, then the rate of change
exceeds 0.4 ◦ C/decade by 2030 to 2040, although there is some
evidence that the very high end of this range may not be consistent
with observed warming to date17 .
Although the rate of change slows under a 4.5 W m−2 policy,
global temperatures still increase throughout the century. Only a
much stronger climate policy, such as a 2.6 W m−2 peak and decline
scenario (Supplementary Fig. 16), can begin to reduce global-mean
temperatures by the end of the century. Even under such strong
climate mitigation, rates of climate change are above historical
values until mid-century. In contrast, a reference scenario with
higher greenhouse gas emissions, such as the RCP8.5, leads to much
higher rates of change throughout the century (Fig. 4).
Like the projections from more complex models, these results do
not capture the recently observed decrease in the rate of warming.
Owing to internal variability, such a temporary slow-down is neither
unusual nor unexpected8,18,19 , and is consistent with our analysis
of historical rates of change where recent warming does not stand
out when examined over 20-year time periods (Supplementary
Fig. 2). The specific mechanisms at work are still a subject
of research20,21 .
Global temperatures, with the impacts of natural variability
subtracted, are close to, or may have already reached, a 0.2 ◦ C/decade
rate of change22 (Supplementary Fig. 15). Continued research on
near-term climate prediction21,23,24 , including resolving the physical
mechanisms behind observed temperature changes over the past
several decades, is needed to improve our understanding of potential
rates of climate change over the next two to four decades. Improved
methodologies for separating the impacts of natural variability
from anthropogenic influences would also aid in the observational
verification of anthropogenic climate changes over time22 . Natural
variability will play a key role in determining when increases in
the rate of change would be observable. Note that both internal
variability and changes in external forcings (solar and volcanic) have

NATURE CLIMATE CHANGE | ADVANCE ONLINE PUBLICATION | www.nature.com/natureclimatechange
© 2015 Macmillan Publishers Limited. All rights reserved.

3

LETTERS

NATURE CLIMATE CHANGE DOI: 10.1038/NCLIMATE2552

been found to play a role in determining 50–100-year trends in the
absence of anthropogenic forcing changes25 .
The large shift over the coming decades relative to the historical
period of both occurrences of high rates of change and the average
rates of change indicates that the world is now entering a regime
where background rates of climate change will be well above
historical averages until at least mid-century. The impacts of an
extended period of increasing rates of change are unclear. Little
research has been conducted specifically on impacts associated with
differing rates of climate change over a timescale of two to four
decades. Specific impacts, including in natural systems, may be
sensitive to changes over different timescales. Those impacts that
are sensitive to changes over 40-year periods would now be on the
cusp of experiencing unprecedented rates of change. The accelerated
rates of change noted here mean that impacts related to rates of
change will intensify over the coming decades. Research on such
impacts, and also on potential adaptation measures, is urgently
needed to guide adaptation. Adaptation to these changes, as well
as detection and attribution of impacts26 , will be necessary as these
rates of change will be sustained for some decades even under
substantial emission mitigation efforts (Figs 4 and Supplementary
Information 16). Mitigation of greenhouse gas emissions would be
needed to limit the duration and magnitude of high rates of change
over the long term.

9. Smith, S. J. et al. Anthropogenic sulfur dioxide emissions: 1850–2005. Atmos.
Chem. Phys. 11, 1101–1116 (2011).
10. Klimont, Z., Smith, S. J. & Cofala, J. The last decade of global anthropogenic
sulfur dioxide: 2000–2011 emissions. Environ. Res. Lett. 8, 014003 (2013).
11. Thomson, A. M. et al. RCP4.5: A pathway for stabilization of radiative forcing
by 2100. Climatic Change 109, 77–94 (2011).
12. Riahi, K. et al. RCP 8.5—A scenario of comparatively high greenhouse gas
emissions. Climatic Change 109, 33–57 (2011).
13. Morice, C. P., Kennedy, J. J., Rayner, N. A. & Jones, P. D. Quantifying
uncertainties in global and regional temperature change using an ensemble of
observational estimates: The HadCRUT4 dataset. J. Geophys. Res. 117,
D08101 (2012).
14. Smith, S. J. & Bond, T. C. Two hundred fifty years of aerosols and climate: The
end of the age of aerosols. Atmos. Chem. Phys. 14, 537–549 (2014).
15. Smith, S. J., Pitcher, H. & Wigley, T. M. L. Future sulfur dioxide emissions.
Climatic Change 73, 267–318 (2005).
16. Van Vuuren, D. P. et al. Temperature increase of 21st century mitigation
scenarios. Proc. Natl Acad. Sci. USA 105, 15258–15262 (2008).
17. Stott, P. et al. The upper end of climate model temperature projections is
inconsistent with past warming. Environ. Res. Lett. 8, 014024 (2013).
18. Meehl, G. A., Arblaster, J. M., Fasullo, J. T., Hu, A. & Trenberth, K. E.
Model-based evidence of deep-ocean heat uptake during surface-temperature
hiatus periods. Nature Clim. Change 1, 360–364 (2011).
19. Easterling, D. R. & Wehner, M. F. Is the climate warming or cooling? Geophys.
Res. Lett. 36, L08706 (2009).
20. Balmaseda, M. A., Trenberth, K. E. & Källén, E. Distinctive climate signals
in reanalysis of global ocean heat content. Geophys. Res. Lett. 40,
1754–1759 (2013).
21. Guemas, V., Doblas-Reyes, F. J., Andreu-Burillo, I. & Asif, M. Retrospective
prediction of the global warming slowdown in the past decade. Nature Clim.
Change 3, 649–653 (2013).
22. Rahmstorf, S., Foster, G. & Cazenave, A. Comparing climate projections to
observations up to 2011. Environ. Res. Lett. 7, 044035 (2012).
23. Meehl, G. A. et al. Decadal prediction—can it be skillful? Bull. Am. Meteorol.
Soc. 90, 1467–1485 (2009).
24. Hawkins, E. & Sutton, R. The potential to narrow uncertainty in regional
climate predictions. Bull. Am. Meteorol. Soc. 90, 1095–1107 (2009).
25. Schurer, A. P., Hegerl, G. C., Mann, M. E., Tett, S. F. B. & Phipps, S. J. Separating
forced from chaotic climate variability over the past millennium. J. Clim. 26,
6954–6973 (2013).
26. Rosenzweig, C. & Neofotis, P. Detection and attribution of anthropogenic
climate change impacts. WIREs Clim. Change 4, 121–150 (2013).
27. Wigley, T. M. L. & Raper, S. C. B. Reasons for larger warming projections in the
IPCC Third Assessment Report. J. Clim. 15, 2945–2952 (2002).
28. Raper, S. C. B. & Cubasch, U. Emulation of the results from a coupled general
circulation model using a simple climate model. Geophys. Res. Lett. 23,
1107–1110 (1996).
29. Raper, S. C. B., Gregory, J. M. & Osborn, T. J. Use of an upwelling-diffusion
energy balance climate model to simulate and diagnose A/OGCM results.
Clim. Dynam. 17, 601–613 (2001).

Methods
CMIP5 data were processed for the first ensemble member of every model for
which the required data was available over the time span of the analysis. Regional
data were averaged over rectangular regions consistent with those from the
PAGES 2k palaeo-climate analysis (see Supplementary Information).
We examine the factors contributing to these twenty-first century changes by
using the MAGICC simple climate model27,28 as implemented in the Global
Change Assessment Model (GCAM; Supplementary Section 5). The MAGICC
model is a four-box energy balance model with an upwelling diffusion
representation of ocean heat transport that has been shown to be able to
reproduce the global-mean temperature results of more complex models29 . The
results shown here incorporate a comprehensive suite of future anthropogenic
forcing agents, including well-mixed greenhouse gases, sulphate and
carbonaceous aerosols, and ozone. Because the model produces trends without
natural variability, realized trends would have a variable component
superimposed on the model results as shown. We limit the MAGICC sensitivity
analysis to two primary variables (climate sensitivity and aerosol forcing) for
clarity and because these are known to be particularly important in this context.
Varying other model assumptions, such as the long-term rate of ocean heat
uptake4 , would widen the range of results.

Received 24 June 2014; accepted 27 January 2015;
published online 9 March 2015

Acknowledgements

References
1. Smith, J. B. et al. Assessing dangerous climate change through an update of the
Intergovernmental Panel on Climate Change (IPCC) ‘‘reasons for concern’’.
Proc. Natl Acad. Sci. USA 106, 4133–4137 (2009).
2. O’Neill, B. C. & Oppenheimer, M. Climate change impacts are sensitive to
the concentration stabilization path. Proc. Natl Acad. Sci. USA 101,
16411–16416 (2004).
3. Taylor, K. E., Stouffer, R. J. & Meehl, G. A. An overview of CMIP5 and the
experiment design. Bull. Am. Meteorol. Soc. 93, 485–498 (2011).
4. Ross, A., Matthews, H. D., Schmittner, A. & Kothavala, Z. Assessing the effects
of ocean diffusivity and climate sensitivity on the rate of global climate change.
Tellus B 64 (2012).
5. Jungclaus, J. H. et al. Climate and carbon-cycle variability over the last
millennium. Clim. Past 6, 723–737 (2010).
6. Pages 2k Consortium, Continental-scale temperature variability during the
past two millennia. Nature Geosci. 6, 339–346 (2013).
7. Tingley, M. P. & Huybers, P. Recent temperature extremes at high
northern latitudes unprecedented in the past 600 years. Nature 496,
201–205 (2013).
8. Collins, M. et al. in Climate Change 2013: The Physical Science Basis
(eds Stocker, T. F. et al.) Ch. 12, 1029–1136 (IPCC, Cambridge Univ.
Press, 2013).
4

The authors are grateful for research support provided by the Integrated Assessment
Research Program in the Office of Science of the US Department of Energy and the
PNNL Global Technology Strategy Program. The Pacific Northwest National Laboratory
is operated for DOE by Battelle Memorial Institute under contract DE-AC05-76RL01830.
We acknowledge the World Climate Research Programme’s Working Group on Coupled
Modeling and the climate modelling groups (Supplementary Table 2) for producing and
making available their model output. The US Department of Energy’s Program for
Climate Model Diagnosis and Intercomparison provides coordinating support for CMIP.
The views and opinions expressed in this paper are those of the authors. The authors
would like to thank J. Dooley and P. Applegate for helpful comments and J. Seibert for
data analysis.

Author contributions
S.J.S., J.E. and K.C. designed research. S.J.S., C.A.H. and A.M. conducted research. All
authors wrote paper.

Additional information
Supplementary information is available in the online version of the paper. Reprints and
permissions information is available online at www.nature.com/reprints.
Correspondence and requests for materials should be addressed to S.J.S.

Competing financial interests
The authors declare no competing financial interests.

NATURE CLIMATE CHANGE | ADVANCE ONLINE PUBLICATION | www.nature.com/natureclimatechange
© 2015 Macmillan Publishers Limited. All rights reserved.


nclimate2552.pdf - page 1/4
nclimate2552.pdf - page 2/4
nclimate2552.pdf - page 3/4
nclimate2552.pdf - page 4/4

Related documents


nclimate2552
climate hypothesis
climate shift hypothesis
ijetr2216
mine wat environ 28 2009 124 135
nature14283 1


Related keywords